Using Task Runner

Task Runner allows you to define and execute a task, for example sending a query to a data archive or running an application (a piece of software available via the VO), as well as analysing the outputs of the task execution. Task Runner Reference Help provides a summary of the main menus common to all uses of Task Runner.

In most cases a Task Runner window will be opened from VOExplorer, when you select a VO service that requires you to build a query to it, or to define input and output parameters to be sent to the service.

Task Runner can also be opened standalone, in which case it allows you to load a previously saved task and execute it.

Examples of Tasks

Queries to data archives

Some data archives allow you to send selective or complex queries. When you select a data archive with such capability (identified by the icon: ) in VOExplorer, a Build ADQL button appears on the left under Actions. This will allow you to open the Query Builder part of Task Runner and define the query to be sent to the data archive.

See this page for help on how to build a query.

Running applications

Many applications that process astronomical data, operate on results of queries, and also models, can be run via the Task Runner. Examples of these are: CrossMatcher, S-Extractor and Solar Movie Maker. In VOExplorer, they are identified by this icon: .

When you select one of these resources in VOExplorer, an Execute Task button appears on the left under Actions. Click on this and the Task Runner interface appears.

Defining Task Runner inputs and outputs

The screenshot below displays the Task Runner window, for an example application.

The Task Runner interface contains dialogue boxes for your Inputs (left pane) and Outputs (middle upper pane).

Inputs

Hovering the mouse over the input and output boxes in Task Runner displays Information (middle lower pane) on each of the input and output parameters. Input parameters can be numbers or strings (in which case type the required values into the boxes) or can be files. The input files need to be located in VOSpace, therefore, if necessary, upload any required input files before running the task - see File Explorer. To point to the input files stored in your VOSpace, click on this icon under Inputs: and point to the required files. The VOSpace file location will then appear as a string in the input box.

In some cases, a given application parameter will only take pre-defined values (eg a string parameter may only accept the values 'Y' or 'N'): in this case a drop-down box appears next to the parameter name and you can select the value from the drop-down box.

If specified in the registration of the application, default values will appear in the input parameter boxes, but these can be changed as required.

Outputs

Output files for the task are by default Stored in Cache (see screenshot above). This is fine for small output files but if you expect large files to be returned, you should save the output into a VOSpace file: under Outputs click on this icon: to select the VOSpace destination for files written by the task.

See the help for individual tools for more information on their inputs and outputs, and see the Query Builder section for querying data sets.

Running the task

To run the task, click on Execute (top right). This should find the first working service but if you want to select the server use the pull-down menu to its right.

Messages reporting the progress of the task will appear in the Execution pane (top right).

Any error messages will also appear here.

Clicking on the blue arrow to the right of the execution report allows you to view a full transcript, or halt the task. If you select Delete task here, this deletes the progress report, not the entire task.

You can change the input and/or output parameters of the task and re-run it. A second execution report will appear above the previous one (as in the above screenshot).

If you want to re-load the parameters of one of your previous runs into the task runnner, click on the blue arrow next to the corresponding run and choose Populate form.

Viewing the results

Once the job has completed an icon with filename will appear in the Execution pane for each of the outputs.

If the output file is in tabular format, you can view it using Topcat (see Helper Applications). After launching Topcat, click on the result file icon and from the Result menu of Task Runner choose Attempt to send table to Topcat.

You can also load an output file into the deafult browser (View) or save it on your local machine (Download).

Saving Tasks

Details of a task (including the information on the application used and the input and output parameters) can be saved into a task document for future use (task documents are saved in xml format). This is done using File then Save. If you close Task Runner without saving details of the tasks run, these will be lost, as will be the record of task executions. Any output files saved in VOSpace will remain accessible.

To rerun a saved task in a new TaskRunner session, start VODesktop and choose New Task Runner. Then, in the File menu, choose Open and load the previously saved task document.

Examples

• Using Task Runner to cross-match two tables

• Using Task Runner to send a query to a database of solar events